Hi Heidi Hello Guillaume, > > I've actualy never thoguht about HTqPCR and ChIP analysis, but that's > solely because I've never received any ChIP-qPCR data myself. I don't see > any a priori reason why you couldn't use the package, at least for QC and > preprocessing. > > Thank for your answer and sorry for the late :) > > Dear Heidi Dvinge > > > > I want to use your package HTqPCR in order to visualize, normalise and > > analyse my rtqPCR data. So before I want to ask you some questions. My > > little dataset is composed of 40 samples. I want to collect data for 8 > > genes including 2 housekeeping. The number of genes is very small and I > to > > normalize with deltaCt methods. Yet my samples are derived from a ChIP > > experiment. So for all my samples I have 2 extracts, one input, the > > control > > that has not been immunoprecipitated, and an immunoprecipitated > > (IP) extract. PCR is performed on this DNA. The problem is ChIP > > experiment contrain us to have a difference of concentration between > Inupt > > and IP, here input is 50 more concentrate than the IP. > > > Just to clarify, what exactly is it you want to test here? Whether there's > a different between Input and IP, or do you have multiple different > treatments, and you want to test whether the IP is more enriched relative > to the Input in some treatment? And with 50x more concentrated, do you > mean the amount of antibody added initially, or the amount of extracted > material? The objective of this experiment is to distinguish sample groups according to their enrichment for a genomic mark on my six target genes. This genes are located next to each other on the studied genome. Previous results of Chip-array indicates a very similar enrichment of this 6 target genes for this genomic mark on the same individual. This enrichment can be strong or weak, and few intermediate results are found. So I try to recover similar results using ChIP-Chip experiment. So I have just one condition/treatment, and two controls in addition to this 6 target genes, a positive and a negative control. For the concentration factor between IP and Input, it concerned extracted material. IP and Input samples have undergone same experimental steps, except imunoprecipitation step for Input. Input correspond to the total DNA extracted material, and IP correspond to the results of immunoprecipitation of the same DNA against my genomic mark. > > So finally I have 2 Ct value for each genes, Input and IP value, and this > > value don't correspond of the same concentration. My question is, do you > > think your tool can be adapted to my case? In the affirmative, do you > have > > any suggestion about what to normalize my data between Input and IP Ct > > value ? > > > Normalising can admittedly be a bit tricky when you have so few genes. In > your samples and treatments, can you assume that the two housekeeping > genes ought to have the same signal in IP and Input when adjusted for > concentration? And/or will the level of your six other genes vary across > samples, or are they all from the same treatment? > After data acquisition, I have my 2 Ct value for each genes, Input and IP. The first data treatment consisted in transformation of Ct values of Input and IP into a unique enrichment value, using following calculation : 2 ^ ( mean(Ct_input) - mean(Ct_IP) ) * 100/dilution_Factor here dilution_Factor = 50 So my dataset is composed of, for each sample, one enrichment value by target genes. This is this dataset I used whith HTqPCR. So I want to compare global enrichment for my genomic mark on my target genes and to cluterised my sample according to this enrichment values. I test HTqPCR for a preliminary analysis whith this data, and I have an other question for the clusters manipulation. In a first time I use plotCtHeatmap() in order to have a representation of the clustering result : >plotCtHeatmap(data, gene.names="", dist="euclidean") So at this step, I want to extract the different samples cluster, but it is not possible with this function. This work can be done by the clusterCt function : >clusterCt(N.data, type="samples", dist="euclidean", n.cluster=...) but I can't see the heatmap corresponding to this clustering, and the clustering using plotCtHeatmap() and clusterCt(type="samples") are not the same (because plotCtHeatmap make clusters by samples and features I think). There is a solution for visualize the heatmap like plotCtHeatmap() and recover cluster lists like clusterCt() ? Thanks for your consideration > > Best, > \Heidi > > > thank you > > > > Guillaume T. > > > > Guillaume Tiberi, ingenieur d'etudes en Bioinformatique > > guillaume.tiberi@inserm.fr > > 04 91 22 33 33 poste 4186 > > Equipe Estelle Duprez > > Centre de Recherche en Cancerologie de Marseille, > > http://crcm.marseille.inserm.fr > > Inserm UMR 891, 27 bd Leï Roure, BP 30059, 13273 Marseille Cedex 09 > > France > > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]